Roles of Chondrocytes in Endochondral Bone Formation and Fracture Repair

Abstract

The formation of the mandibular condylar cartilage (MCC) and its subchondral bone is an important but understudied topic in dental research. The current concept regarding endochondral bone formation postulates that most hypertrophic chondrocytes undergo programmed cell death prior to bone formation. Under this paradigm, the MCC and its underlying bone are thought to result from 2 closely linked but separate processes: chondrogenesis and osteogenesis. However, recent investigations using cell lineage tracing techniques have demonstrated that many, perhaps the majority, of bone cells are derived via direct transformation from chondrocytes. In this review, the authors will briefly discuss the history of this idea and describe recent studies that clearly demonstrate that the direct transformation of chondrocytes into bone cells is common in both long bone and mandibular condyle development and during bone fracture repair. The authors will also provide new evidence of a distinct difference in ossification orientation in the condylar ramus (1 ossification center) versus long bone ossification formation (2 ossification centers). Based on our recent findings and those of other laboratories, we propose a new model that contrasts the mode of bone formation in much of the mandibular ramus (chondrocyte-derived) with intramembranous bone formation of the mandibular body (non-chondrocyte-derived).

Keywords: cartilage; cell transdifferentiation; osteoblast; osteocyte; osteogenesis; temporomandibular joint.

Figure 1.

Cellular co-localization of the chondrocyte-derived tomato marker and a 2.3 Col1a1-GFP osteoblast-specific marker in the growth plate and underlying trabecular bone. (a) Cartoon illustrating the cross that generates triple mice containing Agr-Cre^ERT2, Rosa26^tdTomato, and 2.3Col1a1-GFP, with tamoxifen induction at postnatal week 2 and at week 6 (harvest) for confocal imaging. Confocal images from red channel (b), green channel (c), and merged channels (d), in which the green (arrow) indicates a non-chondrocyte-derived bone cell, red (solid arrowhead) reflects a chondrocyte-derived cell that produces little collagen with no GFP activation, and yellow indicates a chondrocyte-derived bone cell that produces tomato (reflecting aggrecan activity) and type I collagen with GFP activation (open arrowhead; Agr-Cre^ERT2; 2.3Col1a1-GFP).

Figure 2.

Chondrocyte-derived bone cells contribute to condylar neck and mandibular ramus bone formation. (a) Qualitative and quantitative demonstration of the direct transformation of condylar chondrocytes into bone in Col10a1-Cre (activated in hypertrophic chondrocytes at ~E15.5), 2.3Col1a1-GFP (activated in bone cells), and Rosa26^tdTomato triple transgenic mice. The merged P01 (postnatal day 1, left panel) confocal image reveals Col1a1-positive cells (green) in perichondrium and periosteum on the periphery, Col10a1-expressing cells (red) in the top center, and bone cells derived from cartilage or bone marrow (red/yellow; green) in the lower center. The P10 confocal image in the center panel shows a mix of different-colored bone cells in the subchondral bone (white arrows; top center and the enlarged inset on top right), trabecular bone (enlarged insert on left), and periosteum. In the right panel (P21), the low-magnification confocal image illustrates the 3 areas used to quantify bone cell origin: yellow/red marks cells derived from chondrocytes and green marks those originating from non-chondrocyte progenitor cells. Enlarged representative confocal images were used for quantitation in each area (superior, s; mid, m; inferior, i); and the results are presented in the bar graphs in the lower right panel. The figure is modified from Jing et al. 2015. (b) We propose that the mandibular bone is formed by 2 types of bone cells: the chondrocyte-derived bone cell for condylar neck and the upper ramus (endochondral bone origin) and the bone marrow- and periosteum-derived progenitor cell for the surface of the ramus and body of the mandible (intramembranous bone origin).

Figure 3.

Co-localization of the tomato reporter (red) with different immunostained markers (green color) in mice containing Rosa26^tdTomato–Aggrecan Cre^ERT2. The reporter was activated at P14 (postnatal day 14) by tamoxifen. Mice were harvested at P28, and viewed with immunostaining for COL1a1 (early bone marker, a), or DMP1 (late bone marker, b), or COL2 (chondrocyte marker, c); (d) Safranin O-stained section showing the cartilage remnants in the subchondral bone area to trace the fate of chondrocyte-derived bone cells. Ob, osteoblast; Ocy, osteocyte.

Figure 4.

The condyle ossifies at a single ossification center versus 2 ossification centers in the limb. (a) Cartoon illustrating the cross that generates the compound mouse containing Col10a1-Cre and Rosa26^tdTomato. The confocal images show a gradually expanding ossification area at stages E16.5, E18.5 and P01. (b) Cartoon illustrating the cross that generates the compound mouse containing Agr-Cre^ERT2, Rosa26^tdTomato, and 2.3Col1a1-GFP, with tamoxifen induction at postnatal day 14 (P14) and harvested at P28 for confocal imaging. (c) Cartoon illustrating the ossification process in long bone, in which the first ossification center is formed during the embryonic stage and the second ossification center appears after birth, with the growth plate vital for bone growth. (d) Cartoon illustrating the ossification process in the condyle, in which the single ossification center is initiated below and continues to expand during growth. For condylar development, the perichondrium plays a role similar to that of the growth plate.

Figure 5.

Deletion of Bmpr1a (the key receptor for BMP2 and BMP4) in chondrocytes leads to a lack of condylar cartilage and a short ramus. (a) Representative radiographs from the Bmpr1a cKO mice (a crossing of Aggrecan-Cre^ERT2 and Bmpr1a loxP with tamoxifen induction at postnatal day 3 and the animal harvested at day 21) reveal not only a major defect in the condylar cartilage, but a short ramus as well (right panel); and (b) Safranin O-stained images demonstrate a lack of active chondrogenesis, and poorly formed subchondral bone with no cartilage residues in the Bmpr1a cKO mice. These data support a strong dependence on subchondral bone formation for healthy chondrogenesis, in which BMP signaling plays a key role.